In this study a solution procedure is developed for a class of industrial investment planning models which incorporates the following features: economies-of-scale in production, intermediate input-output relationships among production activities, and joint production of different products having common processing requirements (capacity sharing). The model is formulated as a mixed integer programming problem. It is single-period and disregards spatial considerations. The choice is between domestic production and imports (make-buy) to satisfy exogenously stated demands for a given set of interrelated products. 

A two-stage solution procedure was developed and specialized to various specifications of the planning model. At the first stage simple sufficiency conditions for import and for domestic production of a given product are systematically applied in an attempt to reduce the size of the problem. At the second stage (solution stage) an LP-based branch-and-bound (B-B) algorithm is used. 

Data from the mechanical engineering (metal working) sector of the Republic of Korea was used to implement the proposed two-stage solution procedure. The results from 25 test problems generated from the Korean data provide strong evidence of the efficiency of the approach. Moreover, computational experience with the B-B stage alone indicates that very large problems can be efficiently solved without dependence on the success of any form of problem reduction attempt.